reallycurious

My SaaS product extracts data from pdf files. Free trial gets users 10 free extractions. Paying users get 100 jobs. User logs into dashboard, uploads pdf, and downloads the extracted data. That is one job.

While observing 50 users so far I notice that the user engagement seems to terminate after they run the job for the first time. Even if it is successful. Some users run it a few more times but the result is the same, they leave and never come back or they come back run a job and leave again. Another observation is that 30 out of 50 users never created any jobs (they must label the pdf using the web app before extraction can happen, I explain how to do it in the demo video).

Even reaching out to all 50 users asking if I could help them, I got one reply. Here's what one of my user said

"Give that your subscription packages allow customers to run a certain number of jobs, it’d be important to me as a customer to know that I don’t waste jobs trying to figure out what data are captured every time and refining the same job until I get it right. "

Is limiting the free trial to 10 jobs per user be the cause of engagement terminating after running a few jobs? I reached out to the user to see how they would respond to having no limits.

Are people having trouble with just figuring out how to use the tool evidence by a over 60% inactivity? Whats more frustrating is that they do not respond to my email queries where I ask them if they need help.

I would like to study more about user engagements and creating a business model that will get users to engage with the product in order to reach a key activation metric to find value in the product.

Background: My product extracts data from pdf files. Free users get 8 free jobs. Paying users get 40 jobs. Each job is a data extraction from a pdf file.

Problem: User engagement terminates after they run into a bad extraction. On some pdf files it just doesn't work, it returns empty data. Most times it extracts data but not exactly to what the user had in mind (duplicate, mismatching column, unwanted data). Right after this happens, the user will either try a few more times with other pdf files or just give up and never return again. To fix the cases where it returns empty data, it takes me at least 10 hours.

Concern: I can't support myself to work on such fixes and I can't see the number of bad data extraction that needs to be fix decreasing for each user. It's very tough problem to extract data from pdf files, it's never perfect but I was hoping it was good enough for some users. Even a paying user with their number of fix requests, I would have a hard time getting to them all. The bigger problem is that I priced is cheap so I would get more users, turns out this would lead to a far greater number of problems being encountered with the data extraction process.

Search for solution: My immediate thought is to ditch the free plan, increase the product price and somehow deal with the number of fixes being requested and barely breakeven.

I don't know if anyone's been in this situation, it almost seems like the end user cares about the result of the data extraction unlike say a bookkeeping web application that has a much narrow scope of problems expected. Like if a pdf doesn't extract correctly, it's of no value to the user so they will want this to be fixed right? And as they keep encountering more and more fixes I won't be able to keep up....I'm really about call it quits I've worked long and hard on this project and I don't think I can go on.